[00:55] rcn-ee: haha, you just need to load g_ether! [00:55] dammit [00:55] bug 608312 [00:55] Launchpad bug 608312 in linux (Ubuntu) "Usb host mode on OTG doesn't work on Maverick with BeagleBoard (affects: 1) (heat: 8)" [Undecided,New] https://launchpad.net/bugs/608312 [00:55] then the OTG works as expected [00:55] at your config file you have this as built-in [00:55] well, time for some sleep now :-) [00:55] * rsalveti -> out [01:14] good to hear rsalveti (yeah the musb stuff is very fragile, one little thing will kill it.. ;) ) [05:20] long day === bandwidthcrunch is now known as yonathanaw [08:17] morning === XorA|gone is now known as XorA [08:37] morning [09:12] cooloney: I updated the otg bug, take a look at it later [09:17] Morning mythripk [09:17] is there a list of supported omap3/4 machines? [09:18] XorA, not yet since the kernel team is still looking into patces [09:18] *patches [09:18] hehe, suck it and see [09:18] * XorA would love zoom3 and 4430sdp :-D [09:18] XorA, safe bets are panda, beagle C4/XM and blaze [09:19] blaze will need some tinkering with the bootloader on the image [09:19] hey vincent [09:19] i plan to provide a script that replaces u-boot and MLO with the ones you want for a target, the kernels should be generic enough to run on all the platforms === vincent is now known as Guest55889 [09:20] that reminds me, wonder if u-boot for 4430 supports nand yet [09:20] given that there are no devices with NAND, why would anyone enable it ? [09:26] ogra: no devices with nand???? [09:26] * XorA has a deskful of devices with nand [09:26] 4430sdp with NAND ??? [09:27] * XorA thought it did [09:28] * XorA has his hand hacked image on 4430 [09:29] rsalveti: exactly, sorry for not syncing with your [09:29] rsalveti: i tried build in the g_ether into the kernel [09:30] the musb hub will be found [09:30] cooloney: that will work, but I don't know if this is the "right" solution :-) [09:30] don't know if building g_ether as built-in is going to affect something else [09:30] rsalveti: yeah, because, i found there are 2 place to usb_add_hcd(musb_hcd) [09:31] rsalveti: 1 is in host only mode [09:31] yep [09:31] 2 is in the OTG mode and gadget init code [09:31] but the gadget init code will be only called by the gadget upper level class driver when it start [09:32] rsalveti: i tried the compile the musb as host only controller [09:32] rsalveti: it works now [09:32] at least host function [09:32] cooloney: but why do you want as host-only? [09:33] rsalveti: because i wanna test the usb_add_hcd was called in the first method [09:34] rsalveti: when i built-in the g_ether, musb hub was found, but usb host function does not work [09:34] cooloney: works when you put the otg cable, at least for me [09:35] cooloney: both gadget and host mode starts to work fine [09:38] rsalveti: ok, got it. [09:38] cooloney: if you just change the g_ether as built-in everything should work just fine [09:38] rsalveti: so for summary, if we wanna use the host mode of musb, we need to modprobe g_ether, right? [09:39] cooloney: then we need to check if the same modification works for lucid [09:39] cooloney: currently, yes [09:39] rsalveti: yeah, yeah, that's what i want to do. [09:39] rsalveti: but it failed, it makes me frustrated [09:39] cooloney: because loading g_ether will initialize the gadget and host functionality of the otg chip [09:40] rsalveti: exactly, [09:41] * cooloney nods to rsalveti and shakes hands with rsalveti [09:41] rsalveti: ok, let me try again. [09:41] cooloney: ok :-) [09:41] * rsalveti still needs to get a working lucid environment [10:16] Is rootstock about to make a lucid rootfs? === User1 is now known as Jameswstubbs === persia` is now known as persia [10:17] ogra ? [10:17] able* [10:20] It keeps failing at the second stage [10:25] Jameswstubbs: yes, you can create a lucid rootfs with it [10:25] Jameswstubbs: what kind of failure? qemu seg fault or just got stuck while doing the second stage? [10:29] Stuck during second stage [10:30] Well, when starting second stage [10:30] You might be interested in this actually, ixproject.org :) [10:31] My problem is my karmic rootfs uses an old synaptic module [10:31] I tried to add the lucid repo to the karmic apt and tried to update the module, but got a fatal kernel panic [10:31] I've decided to maybe try lucid, should still be compatible with armv6 shouldn't it? [10:32] Jameswstubbs: No, Lucid and future releases are armv7. [10:34] rsalveti: good news, i build in g_ether this time, it works [10:35] cooloney: cool [10:35] Jameswstubbs: please paste the rootstock log at the paste.ubuntu.com and send me the link [10:35] rsalveti: but it looks like a design issue, i think the usb_add_hcd should be called in the early init code, not in the late init code in gadget [10:36] Jameswstubbs: it all depends on how you're running it, as we have these 2 known issues [10:36] cooloney: yep, that's what I think [10:36] rsalveti: it is stupid to get host function by insmod a gadget class driver or build in it [10:36] cooloney: haha, for sure :-) [10:36] rsalveti: give me sometime to write a patch [10:36] cooloney: that's why I was so mad yesterday [10:36] rsalveti: haha, yeah, i was totally frustrated by that as you did [10:37] cooloney: cool, take your time to write the proper patch and send it to linux-omap [10:37] most of the musb maintainers are looking at that list [10:37] Ok, isn't armv7 backwards compatible with v6? [10:37] I'm installing a netinst of ubuntu lucid now [10:38] I'll have to rerun rootstock then paste the log [10:40] Jameswstubbs: armv7 is backards compatible hardware wise. But don't expect an armv6 system to run armv7 software. [10:40] Hm, if lucid is indeed completely unsuitable for armv6, how would I go about upgrading the xserver-xorg-input-synaptics. [10:40] If I find the source code for the latest version [10:40] And compile from within my karmic rootfs in qemu, would that work? [10:41] Are the modules architecture dependant ? [10:42] Jameswstubbs: What exactly are you trying to do? If you are running Karmic on armv7 hardware, then it is possible to upgrade to Lucid. But if your hardware is armv6 only, it won't work. Like trying to run x86_64 software on a Pentium. [10:43] I'm porting ubuntu to the iPhone, I have a karmic rootfs fully up and running with wifi, xorg and xfce [10:43] ixproject.org if you want a screenshot [10:43] Having a look at my xorg logs, the iPhone multitouch screen uses a synaptic module [10:44] But then says it cannot find a support touchpad [10:45] Jameswstubbs: if you build it at Karmic it should work fine [10:45] pastie.org/1054947 from line 192 down [10:46] it's just that Lucid compiles everything for armv7 by default [10:46] I'm a little familiar with ixproject.org. Unfortunately, the hardware just won't run software compiled for the new architecture. You can pull the source for the latest archive and rebuild it, but our stack is designed for armv7 starting with Lucid. [10:46] I only started ixproject last week :S Maybe familiar with idroidproject [10:46] GrueMaster: I think that's what I'll do [10:46] Jameswstubbs: you could get stuff from debian if you like, it's armv5 compatible [10:46] I'll hunt down the source and compile from within qemu [10:47] The iPhone is armv6 [10:47] Best of luck on your endeavours. [10:47] I've had a debian armel rootfs running happily [10:47] I simply like Ubuntu :) [10:47] Debian is still armv4 iirc. [10:47] Thanks GrueMaster, I've needed luck up to now and I'll keep needing it [10:47] heh. :) [10:48] GrueMaster: I simply used multistrap with debian squeeze [10:48] Worked a charm [10:48] cool [10:48] I have someone working on USB drivers aswell [10:49] Should hopefully have USB host mode running soon [10:50] Jameswstubbs: awesome, love your work, [10:50] Thank you :) [10:50] although i've sold my iphone and switch to N1 [10:51] I hope to have touch sorted within a week and a release should be out soon [10:51] but if Ubuntu is ready for iphone, i will grab my girl friend's iphone and force her switch to Ubuntu for iphone [10:51] I don't blame you N1 is an awesome phone, only reason I'm still using iPhone is becasue Im stuck in contract for another year [10:51] cooloney: I wouldn't recommend it immediatley, it wont be ready for prime time [10:52] It needs stripping down and made more mobile friendly [10:52] Im think of using thje matchbox window manager [10:52] Jameswstubbs: yeah, fully understand. i am waiting for iphone4 too, and practicing the holding style suggested by Jobs [10:52] Also power management needs to be implemented, the battery wears down in less than 2 hours :/ [10:52] Jameswstubbs: so how about the kernel in your Ubuntu iphone image [10:53] Jameswstubbs: is the kernel from planetbeing? [10:53] Yep, a couple of small changes concerning USB gadget [10:53] But essentially the stock kernel [10:54] The kernel is not in the image, it's stored on the "host" partition and chainloads the rootfs using initrd [10:54] Havn't been able to get initramfs working with the kernel [10:54] Not sure why [10:54] I have a 40 second boot anyway, so It's not a massive concern [10:54] Jameswstubbs: yeah, [10:56] Jameswstubbs: so is the Ubuntu image only for 3GS or older machine [10:56] We don't have an iboot exploit for the 3gs yet, so development hasn't even started hardware wise [10:56] This is for the 2g and 3g [10:58] I don't own the 2g so havn't been able to test :-/ the only difference is in how the firmware is loaded by the initrd. [11:07] I'm off, thanks for the help :) [11:54] mythripk: I'm back [12:05] apw: bug 608674 and bug 603087 [12:05] Launchpad bug 608674 in linux (Ubuntu) "provide packaging rules in linux-source binary package (affects: 1) (heat: 8)" [Undecided,New] https://launchpad.net/bugs/608674 [12:05] Launchpad bug 603087 in linux (Ubuntu) "Allow to build just linux-libc-dev (affects: 1) (heat: 270)" [Wishlist,In progress] https://launchpad.net/bugs/603087 === robclark1 is now known as robclark [13:22] rsalveti, dd if=/dev/zero of=livecd.${FSS}.ext2 bs=1024 count=0 seek=$size [13:22] rsalveti, size=$(($(du -ks ${ROOT} | cut -f1) + (10240))) [13:29] ogra: What monitor do you have? The one that _just works_ [13:34] lag, samsung syncmaster T240 [13:47] ogra: Hey, did you folks change from .bz2 to .gz recently for omap images? [13:49] vstehle: robclark: http://maemo.gitorious.org/fremantle-hildon-desktop/clutter_0_8 [13:49] the call is EGL_NOKIA_texture_from_pixmap [13:49] at clutter/eglx/clutter-eglx-texture-pixmap.c [13:52] rsalveti: thx [13:59] lool, yes [13:59] lool, bz2 is BAD with rsync, .gz works with rsyncing images [14:35] hello everybody [14:35] now I'm definitely able to run the image from the rootstock! [14:36] but I've some problem with udev and mmcblock and mtdblock... udev does not create the device for the partition automatically... [14:37] anybody has a similar problem? [15:11] lool, yes, bz2 wasnt rsyncable, gzip has the special flag to make the files usable with rsync [15:50] hey all.. I am trying to bring up ubuntu on Blaze (OMAP4) using http://www.omappedia.org/wiki/Get_started_with_ubuntu_on_omap4 but it is locking up at boot up has any one seen this? === orbarron1 is now known as orbarron [15:53] ogra: But doesn't it contain a filesystem? [15:53] lool, sure, whats the issue with that ? [15:54] ogra: Well filesystems aren't very rsync friendly across runs [15:54] I have the issue with cloud images [15:55] they almost don't compress with rsync, despite being built with gzip --rsyncable [15:57] i didnt test to zsync yet [15:57] since we only have one image thats gz yet [16:10] we'll see if there is improvement once i could test :) [16:23] guys, anybody have problem with ubuntu lucid and udev for the automatic creation of /dev/mmcblk0p* and /dev/mtdblock* devices? [16:23] nope, works fine here [16:24] Hey guys... just to let you know OMAPpedia is having a tutorial hour next week covering Uboot and Ubuntu. Visit: http://www.omappedia.com/wiki/Main_Page. The time may be a bit late for some of you but we are trying to find a good time for everyone. orga will be a guest speaker for this session. If anyone is interested in doing a session in Aug. please ping me :D [16:25] I've created a lucid minimal image with rootstock but when the system starts, I don't see any /dev/mmcblk0p* and /dev/mtdblock* [16:25] furibondox: are you able to see the other devices that are created by udev? [16:26] yes [16:26] furibondox: what kernel are you using? [16:26] 2.6.29 [16:26] furibondox: from where? [16:26] the same kernel works fine with a jaunty distribution created by my own [16:26] from TI PSP [16:27] I think the problem is with the udev and lucid [16:28] http://paste.ubuntu.com/467568/ [16:28] as you can see the partitions are present... but udev does not create the device! [16:29] furibondox: could be a bug with udev with this kernel [16:30] may be... [16:30] this is what I see: http://paste.ubuntu.com/467569/ [16:31] furibondox: you could try to debug udev, to see what it detects normally in your case === fta__ is now known as fta === XorA is now known as XorA|gone === JaMa is now known as JaMa|Away === JaMa|Away is now known as JaMa [21:03] hi, I am trying to install ubuntu karmic on bb using the method specified in http://elinux.org/BeagleBoardUbuntuKarmic , it boots the kernel,after login and password I get stucked with a prompt [21:04] there is any command to install from there? [21:18] lem_, personally I'm using linaro's alpha2 on my bb - https://wiki.linaro.org/Source/ImageInstallation [21:22] tgall_foo: hardly on-topic in this channel. [21:28] thanks for the idea, I will try that [21:33] ojn: given the amount of maverick packages in it, it's fairly on topic ;-) [23:00] ojn: Hey, just curious, are you doing some upstream kernel OMAP work? are you doing that on top of Ubuntu? [23:00] lool: I'm not doing much omap work these days. I've spent more time with tegra lately. And no, given my new project and employer, I am not doing much on top of ubuntu. :) [23:01] omap is still around as a side project, but I haven't spent any serious cycles on it lately [23:01] ojn: Oh, which employer is that? :) [23:01] lool: I'll leave that as an exercise for the reader. :)